kinetic study of hydrogen sul de absorption in aqueous ... · water. chlorine is a top-grade...

Kinetic study of hydrogen sulfide absorption in aqueous

chlorine solution

Jean-Baptiste Vilmain, Valerie Courousse, Pierre-Francois Biard, Mohamed

Azizi, Annabelle Couvert

To cite this version:

Jean-Baptiste Vilmain, Valerie Courousse, Pierre-Francois Biard, Mohamed Azizi, AnnabelleCouvert. Kinetic study of hydrogen sulfide absorption in aqueous chlorine solu-tion. Chemical Engineering Research and Design, Elsevier, 2014, 92 (2), pp.191-204.<10.1016/j.cherd.2013.07.015>. <hal-00875684>

HAL Id: hal-00875684

https://hal-univ-rennes1.archives-ouvertes.fr/hal-00875684

Submitted on 14 Nov 2013

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinee au depot et a la diffusion de documentsscientifiques de niveau recherche, publies ou non,emanant des etablissements d’enseignement et derecherche francais ou etrangers, des laboratoirespublics ou prives.

https://hal.archives-ouvertes.fr

https://hal-univ-rennes1.archives-ouvertes.fr/hal-00875684

Kinetic study of hydrogen sulfide

absorption in aqueous chlorine solution

Jean-Baptiste VILMAINa,c

, Valérie COUROUSSEa,c

, Pierre-François BIARDa,b,c*

, Mohamed AZIZIa,c

,

Annabelle COUVERTa,c

aEcole Nationale Supérieure de Chimie de Rennes, CNRS, UMR 6226, 11 Allée de Beaulieu, CS 50837,

35708 Rennes Cedex 7, France

bUniversité de Rennes 1, CNRS, UMR 6226, 3 rue du Clos-Courtel, BP 90433, 35704 Rennes Cedex 7,

France

cUniversité européenne de Bretagne, 5 boulevard Laënnec, 35000 Rennes, France

Graphical abstract

* Corresponding author: Pierre-François BIARD ; Tel: +33 2 23 23 81 57

Email-address: [email protected]

LiquidGas Gas

Liquid

Abstract

Hydrogen sulfide (H2S) is currently removed from gaseous effluents by chemical scrubbing using

water. Chlorine is a top-grade oxidant, reacting with H2S with a fast kinetic rate and enhancing its

mass transfer rate. To design, optimize and scale-up scrubbers, knowledge of the reaction kinetics

and mechanism is requested. This study investigates the H2S oxidation rate by reactive absorption in

a mechanically agitated gas-liquid reactor. Mass transfer (gas and liquid sides mass transfer

coefficients) and hydrodynamic (interfacial area) performances of the gas-liquid reactor were

measured using appropriated physical or chemical absorption methods. The accuracy of these

parameters was checked by modeling the H2S absorption in water without oxidant. A sensitivity

analysis confirmed the robustness of the model. Finally, reactive absorption of H2S in chlorine

solution for acidic or circumneutral pH allowed to investigate the kinetics of reaction. The overall

oxidation mechanism could be described assuming that H2S is oxidized irreversibly by both

hypochlorite anion ClO- (k = 6.75 10

6 L mol

-1 s

-1) and hypochlorous acid ClOH (k = 1.62 10

5 L mol

-1 s

-1).

Keywords

Hydrogen sulfide, chlorine, kinetics, mass transfer, absorption, scrubber

Highlights

• Kinetics of H2S oxidation by chlorine is investigated using a well agitated reactor

• Mass transfer and hydrodynamics performances of the reactor are determined

• Sensitivity analysis is performed to assess the robustness of the modeling

• H2S reacts irreversibly with hypochlorous acid and hypochlorite anion

• Kinetic constants are determined

1 Introduction

Hydrogen sulfide (H2S) is a major toxic compound involved in odor emissions of many industries such

as waste water treatment, pulp industry, etc. (Gostelow et al., 2001; Kangas et al., 1984; Rappert and

Müller, 2005). Proving high and reliable efficiency, chemical wet scrubbing is the most widely used

process to prevent H2S emissions (Busca and Chiara, 2003; Smet and Langenhove, 1998; Smet et al.,

1998). This technique involves H2S mass transfer in an aqueous phase using a gas-liquid contactor.

Since H2S is poorly soluble in water, to increase its removal efficiency, aqueous oxidant solutions at

basic pH are often used (Biard et al., 2009; Bonnin, 1991; Chen et al., 2001; Kerc and Olmez, 2010; Le

Sauze et al., 1991). Other processes are currently used to remove H2S : biofiltration, absorption in

alkanolamine solutions, absorption using the Claus process or catalysts to recover the elemental

sulfur, etc. (Busca and Chiara, 2003; Kohl and Nielsen, 1997).

Due to a strong reactivity with H2S, chlorine is a top-grade oxidant for this application and is currently

used, especially in waste water treatment plants, in rendering and composting facilities, etc. The

process is operated semi-continuously (Biard et al., 2010). The scrubbing liquid is stored in a tank

located at the bottom of the scrubber and is recirculated. pH and chlorine concentration are

regulated. This scrubbing liquid is just drained when the salt accumulation due to H2S oxidation into

sulfate anions becomes too high.

Chlorine speciation in water is rather complex. Depending on the pH, the ionic strength and the

chloride concentration, chlorine can be present in water mainly as aqueous chlorine Cl2,

hypochlorous acid ClOH and hypochlorite ions ClO- (Deborde and Von Gunten, 2008):

+ -

2 2Cl + H O ClOH + H + Cl� Eq. 1

2

+ --4 2 -2

Cl

2

[ClOH][H ][Cl ]K = = 3.8×10 mol L at 293 K and a negligible ionic strength

[Cl ] Eq. 2

- +- +

A,1 A,1

[ClO ][H ]ClOH ClO + H K = with pK = 7.58 at 293 K

[ClOH]�

Eq. 3

Cl2 exists only at acidic pH (pH < 5-6) and its coexistence domain increases with the total chloride

concentration. Some additional species such as Cl3-, Cl2O can be observed. However, at a low chloride

concentration, they can be neglected (Deborde and Von Gunten, 2008). With a higher oxidation

potential, ClOH currently reacts faster than ClO-. These observations emphasize that oxidation

kinetics with chlorine are pH dependant.

An excess of chlorine (at least 4 mol of chlorine for 1 mol of H2S) can mineralize H2S (oxidation

number = -2) in sulfate ion (oxidation number = +6) according to the following global mechanism at

alkaline pH:

complete oxidation

2 2 4 2H S + 4NaOCl + 2NaOH Na SO + 4NaCl + 2H O→ Eq. 4

The first by-product formed after H2S oxidation is elemental sulfur which can be further oxidized in

sulfite SO32-

and sulfate SO42-

(Choppin and Faulkenberry, 1937). In the whole process, oxidation of

elemental sulfur into sulfite is considered as the rate limiting step, especially at pH close to the

neutrality where colloidal sulfur accumulation can be observed in chemical scrubbing processes

(Biard et al., 2009; Bonnin, 1991).

H2S is a weak diacid in equilibrium with its conjugated base, the hydrosulfide anion HS- (pKA,2 = 7.07

at 293°K (Roustan, 2003). The sulfide anion S2-

is present only at very basic pH and can be neglected

(pKA,3 = 13.94 at 293°K) at pH used in chemical scrubbing (pH ≤ 11-12). During H2S chemical scrubbing

in chlorine solutions, both H2S dissociation in HS- and oxidation decrease the H2S bulk concentration,

which maintains a maximal and constant mass transfer driving force. Moreover, since the kinetics of

these reactions are really fast, mass transfer enhancement is currently observed, which leads to a

large removal efficiency (Biard et al., 2010). Indeed, mass transfer enhancement due to these

reactions enhances the absorption rate by decreasing the mass transfer resistance in the liquid film

(Roustan, 2003). By this way, the whole resistance could be located in the gas phase. Mass transfer

enhancement calculation is rather complicated and depends on several parameters: nature

(reversible, irreversible) and number of reactions involved, reaction(s) kinetics, reaction(s)

stoichiometry, concentrations and diffusion coefficients (van Swaaij and Versteeg, 1992).

Within the current operating conditions, which insure an excess of chlorine (several hundred mg L-1

are usually set), H2S and HS- cannot be measured in the liquid bulk since they are quickly oxidized in

the liquid film at the vicinity of the gas-liquid interface. From wet scrubbing experiments in packed

column, Bonnin (1991) determined an apparent kinetic constant between H2S and ClOH of 1.8×108 L

mol-1

s-1

for pH ranging from 9 to 11 at 293 K. However, this determination was based on wrong

assumptions (Biard et al., 2010). Therefore, the kinetic rate of H2S oxidation by chlorine is poorly

known whereas it would enable better gas-liquid contactor scale-up, design and optimization.

To answer to this issue, the purpose of this study is to determine the H2S oxidation kinetics at 293 K

with ClOH and ClO-. Only a few techniques like stopped flow or competitive methods or reactive

absorption can be used to determine high kinetic constants (Beltrán, 2004). In this particular case,

reactive absorption seems to be appropriated (Vaidya and Kenig, 2007b). This technique has been

successfully applied in the past to determine kinetic constants in the fields of ozonation, chlorine

hydrolysis, flue desulfurization or CO2 capture in amine solutions (Aieta and Roberts, 1986; Beltrán,

2004; Ebrahimi et al., 2003; Jia et al., 2010; Jing et al., 2012; Kucka et al., 2002; Sema et al., 2012;

Sotelo et al., 1991; Vaidya and Kenig, 2007a, 2009).

During H2S absorption, H2S removal efficiency is governed by the mass transfer rate which depends

on the hydrodynamic and mass transfer performances of the reactor, gas and liquid flow patterns

and H2S oxidation kinetic rate. If all these parameters are controlled, the kinetics can be investigated.

To simplify the liquid flow pattern and to insure a homogeneous liquid phase, a gas-liquid contactor

agitated by a Rushton turbine has been selected. First of all, the mass transfer and hydrodynamic

parameters of the laboratory tank must be determined in controlled conditions (gas and liquid flow

rates, pressure, rotation speed). These parameters are determined by physical absorption of oxygen

and by chemical absorption of ozone and H2S in specific mass transfer regimes. Then, the kinetic

constant at 293 K can be deduced during H2S absorption in chlorine solutions set at various

concentrations and pH. Chlorine concentration must be optimized to find a compromise between a

too low kinetic rate on the one hand and a too large kinetic rate on the other hand to enable an

accurate enhancement factor determination. pH will be in the range 5-7.5 to assess the influence of

the chlorine speciation and to limit the potential HS- formation. The next section reviews the mass

transfer theory which is required for the gas-liquid contactor characterization and the kinetic

constant determination.

2 Mass transfer modeling

Gas-liquid mass transfer can be described according to various theories. Usually, the double film

theory (explicative scheme in the supplementary material) is applied in steady state (Whitman,

1923). The molar flux transferred of any species A (dJ in mol m-2

s-1

) per square meter of gas-liquid

area can be written (Perry and Green, 1997):

( ) ( )*d

Eq

L A A L A AJ Ek C C K C C= − = −

Eq. 5

E is the so-called enhancement factor (no unit) which is the ratio of the absorption rate with and

without the reaction for an identical difference*

-A A

C C . Other definitions of the enhancement factor

can be found in the literature. However, this one remains the most encountered. CA, *,

Eq

A AC C are the

solute concentration respectively in the bulk (mol m-3

), at the gas-liquid interface (superscript *) and

in equilibrium (superscript Eq) with the gas phase concentrationG

AC deduced from the Henry’s law:

G

Eq A A

A

A A

p R T CC

H H

× ×= = Eq. 6

pA is the partial pressure (Pa), HA the Henry’s law constant (Pa m3 mol

-1). kL and KL (Eq. 4) are

respectively the local and overall liquid side mass transfer coefficients (m s-1

) which are linked

together through the following relation:

1 1

L L A G

RT

K Ek H k= +

Eq. 7

kG is the local gas side mass transfer coefficient. kL and kG order of magnitude are respectively 0.5-5

10-4

m s-1

and 0.5-5 10-2

m s-1

for most gas-liquid contactors (Roustan, 2003). The liquid-phase relative

mass transfer resistance RL , which represents the ratio of the resistance located in the liquid phase

divided by the total resistance, is calculated by the following equation (Hoffmann et al., 2007; Rejl et

al., 2009):

1

1 and 1L

L G L

A G

RTEkR R R

H k

−

= + = −

Eq. 8

In addition to the interface mass transfer and chemical kinetics, the performance of a gas-liquid

contactor depends on the extent of mixing in both the phases (Joshi et al., 1982). Assuming that the

liquid flow is perfectly mixed, i.e. complete back-mixing, the solute bulk concentration is uniform in

the whole reactor which possesses an effective volume V (gas and liquid volumes sum). The extent of

mixing for the gas phase depends on the impeller speed and the gas flow rate through the gas hold-

up (Lee and Pang Tsui, 1999). At a moderate impeller speed, the power input is not sufficient to

disperse the gas in the part of the reactor located below the sparger. In this case, the extent of back-

mixing is low and the flow can be considered as a plug flow. The role of the agitation is to break

bubbles to maintain a high interfacial area. Assuming a plug flow for the gas and a perfectly mixed

liquid, the absorption rate (N in mol s-1

) can be calculated by the following equation:

( ) ( ) ( ), ,

,

,

ln

Eq Eq

A i A A o AEq

L A A L Eq

A i A

Eq

A o A

C C C CN K a V C C K a V

C C

C C

− − −= ° − = °

−−

Eq. 9

Subscripts i and o refer to the reactor inlet and outlet. a° is the gas-liquid interfacial area (m2 m

-3)

relative to the effective volume V. The mass balance enables to write:

( ), ,

G G

G A i A oN Q C C= −

Eq. 10

QG is the gas volume flow rates (m3 s

-1). From Eqs. 6-10:

,

,

ln1

G

A i

A

A

LG

A o A G A G

AL A GA

RTCC

H VRT a VRTK a

RTRTC H Q H QC

Ek H kH

−°= ° =

+− Eq. 11

When the reaction rate is significantly larger than the mass transfer rate, the solute is entirely

consumed in the liquid film and CA tends toward 0. In the other cases, CA depends on the liquid hold-

up, liquid residence time and on the kinetic and mass transfer rates (Lee and Pang Tsui, 1999). Except

for two limit cases (slow and instantaneous reactions in the liquid film), E depends on the reaction

kinetics. Therefore, its determination in controlled and appropriated operating conditions could

enable kinetics determination. kL and kG are mainly influenced by the properties and hydrodynamics

of respectively the liquid and the gas phases. The interfacial area depends on the gas hold-up and

bubbles size, which are determined by the flow field, the gas flow-rate, the power input and the gas

dispersion characteristic of the impeller (RESI, 2013). Except the gas side mass transfer coefficient

which has never been measured for a gas-liquid agitated tank, they can be estimated using semi-

empirical correlations (Calderbank, 1958; Calderbank, 1959; Joshi et al., 1982; Lee and Pang Tsui,

1999). In this study, for a better accuracy, they will be determined prior to E measurements.

3 Material and methods

Different chemical systems were implemented to determine a° (ozone absorption in a solution of

hydrogen peroxide), kL (physical absorption of pure oxygen), kG (chemical absorption of H2S in highly

concentrated chlorine alkaline solution) and E (H2S chemical absorption in slightly acidic or neutral

chlorine solution) (Table 1). H2S absorption in pure water was also investigated to assess the accuracy

of a°, kL and kG determination. The parameters are deduced from the removal efficiency (Eq. 12) of

each absorbed species:

−= , ,

,

G G

A i A o

A G

A i

C CEff

C Eq. 12

Table 1: Synthesis of the physical and chemical operating conditions applied.

Parameter Value Experimental

uncertainty

Physical operating conditions

Temperature T 293 K ± 1 K

Effective volume V 1.65 L ± 0.05 L (3%)

Impeller speed 250 rpm

Liquid flow rate QL 7-30 L h-1

3.0%

Gas flow rate QG 400-750 NL h-1

3.1%

Chemical operating conditions

a° measurement

O3 absorption in H2O2

solution

Inlet H2O2 concentration 1173-2720 mg L-1

2.1%

pH 9.9-10.15 ± 0.1 unit

Inlet O3 concentration 35-60 g Nm-3

2%

kL measurement

O2 physical absorption

pH 6.5 ± 0.1 unit

Inlet O2 concentration Pure oxygen

kG measurement

H2S absorption in ClOH

solution

pH 10.4-10.8 ± 0.1 unit

Inlet H2S concentration 110-135 mg Nm

-

3

10%

Inlet chlorine concentration 1 g.L-1

2.1%

Absorption of H2S in

water

pH 8-8.1 ± 0.1 unit

Inlet H2S concentration 50-65 mg Nm-3

10%

Absorption of H2S in

chlorine solution

pH 5-7.5 ± 0.1 unit

Inlet H2S concentration 50-150 mg Nm-3

10%

Inlet chlorine concentration 50-350 mg L-1

2.1%

3.1 Presentation of the experimental set-up

The laboratory-scale glass gas-liquid contactor (1 in Fig. 1) is a nominal 2 liters tank mechanically

agitated by a Rushton turbine (impeller diameter of 4.5 cm and vessel diameter of 11.0 cm). The

liquid phase is prepared in a 50 L Nalgene® storage tank (2) using demineralized water. Concentrated

NaOH or H3PO4 solutions are added to set the pH. Depending on the chemical system implemented,

chlorine or hydrogen peroxide can be also added to set a given oxidant concentration. The reactor is

fed and drained by an inlet (3) and an outlet centrifugal pumps (4). Flow rates are controlled by

means of valves and measured by two flowmeters (3’,4’). The contactor temperature is controlled at

20 ± 1 °C with an isothermal liquid circulating in the reactor jacket (5). The pH of the liquid and the

temperature are measured on-line using a combined pH-probe (6). Fresh air is introduced using a gas

pump (7). Then a gas flow of N2-H2S (8) is mixed with the pumped air thanks to a static mixer (10).

These two flows are controlled by means of valves and measured by float-type flowmeters (7’,8’)

(Sho Rate R6-15-C and R6-15-D) and dispersed in the reactor by a titanium sparger (11). Before

starting the experiments, the gas-liquid contactor is by-passed to analyze the gas inlet at steady-

state. For the interfacial area determination, the feed system is replaced by an ozone generator fed

with oxygen. In this case, the ozone flow rate is controlled by a flowmeter in Teflon®. During the

experiments, a fraction of the gas is by-passed to analyze the outlet at steady-state (12). Before

being released in the atmosphere, residual O3 and H2S in the gas phase were quenched using

respectively an ozone destructor and by bubbling in a concentrated basic chlorine solution (13). The

gas pressure is measured twice, at the flowmeter outlet to correct the flow rate and in the reactor

headspace, using manometers.

Each float-type flowmeter has been calibrated under controlled pressure and temperature. The read

gas flow-rate must be corrected taking into account pressure and temperature variability using the

following relations (Perry and Green, 1997):

( ) ( ) ( ) ( )-1 -1 -1 -1. . . .

, ,

. .

NL.h NL.h and L.h L.h calib calib calib calib

G G read G G read

calib calib

PT M TP MQ Q Q Q

P TM T PM= =

Eq. 13

The subscripts “calib” refers to the conditions used during the calibration. Values of the physico-

chemical properties and a list of the suppliers and references of the chemical compounds and

equipments used are respectively presented in Tables A.1 and A.2.

Figu

3.2 Analysis

Chlorine and hydrogen peroxide

potassium was oxidized in diode a

was titrated by sodium thiosulfat

analysis to catalyze the I2 formation

H2S in the gas phase was bubbled

temperature and pressure) in a 90

then, oxidized into sulfate ions. The

conductivity detection. The H2S co

the volume of gas trapped.

O3 was analyzed in the gas phase u

were not necessary since they w

therefore could not be detected.

used to measure dissolved oxygen

Page 10

Figure 1: Scheme of the experimental set-up.

eroxide were analyzed by the iodometric method. An

diode at acidic pH (use of HCl for H2O2 and of CH3COO

hiosulfate. Ammonium heptamolybdate at 5 g L-1

wa

ormation.

bubbled continuously at a controlled flow rate of 50

) in a 90 mL concentrated chlorine solution (500 mg

Then the sulfates formed were analyzed by ion chr

S concentration was deduced from the amount of

phase using an UV ozone analyzer. Measurements of dis

they were rapidly oxidized in the chemical system

tected. For kG determination, a membrane covered ga

oxygen concentration.

thod. An excess of iodide

COOH for ClOH) which

was added for H2O2

te of 50 L h-1

(at ambient

00 mg L-1

) at pH = 9, and

y ion chromatography with

ount of sulfate anion and

nts of dissolved O3 and H2S

system implemented and

vered galvanic sensor was

4 Results and discussion

4.1 Hydrodynamics of the gas-liquid contactor

The assumption of the complete liquid back mixing (perfectly mixed liquid phase) has been checked

by Residence Time Distribution (RTD) experiments for various gas (0 to 800 NL h-1

) and liquid (20 to

80 L h-1

) flow rates and impeller speed (100 to 800 rpm) using fluorescein as a tracer. It showed that

the liquid was perfectly mixed while the impeller speed is larger than 100 rpm whatever the flow rate

in the range implemented.

The gas flow pattern cannot be determined by RTD measurement due to a too low residence time

(few seconds) which is not compatible with usual sensor time response. However, we can assume

that the gas back mixing is almost negligible since the impeller speed is lower than the minimal speed

necessary to disperse the gas below the sparger. Indeed, according to the relations cited by Lee and

Pang Tsui (1999), a minimal speed of 800-1000 rpm is expected for the gas flow rate range tested.

This was confirmed by visual observation. Since the gas loading was rather low, the elevation of the

liquid level after starting the gas dispersion was not significant confirming a gas hold-up lower than

1.5%.

The interfacial area a° can be determined in the pseudo first-order chemical absorption regime

(Hoffmann et al., 2007; Rejl et al., 2009). Chemical absorption of ozone in an aqueous solution doped

with hydroperoxyde anions HO2- can be satisfactorily used, especially considering that ozone is poorly

soluble in water (HA is large) which allows to neglect the gas phase resistance (RG →0 and kL ≈ KL).

HO2-, the conjugated base of H2O2, initiates the O3 decomposition according to the following

bimolecular irreversible reaction (Staehelin and Hoigne, 1982):

9 3 -1 -1-O /HO3 2

k = 2.8 10 m mol s- -

3 2 3 2O + HO O + HO °→ Eq. 14

If H2O2 is introduced in large amount and at a basic pH, the ozone consumption is fast in the liquid

film which means that O3 does not reach the liquid bulk. Consequently, O3 is entirely consumed by

HO2- and not by any parasite compound or radical formed after its decomposition with the net

consumption of one mol of H2O2 for one mol of O3 transferred (Biard, 2009). Therefore, according to

Eq. 11:

= °3

3 3

,

,

ln

G

O i

LG

O o O G

C VRTEk a

C H Q Eq. 15

Moreover, since the H2O2 deprotonation into HO2- is really faster than ozone consumption by HO2

-

and since 2 2 3

*

H O OC >> C , the HO2- concentration remains constant in the diffusional film (pseudo first-

order absorption regime). In this case (Hikita and Asai, 1964):

( )tanh

HaE

Ha=

Eq. 16

Ha (Hatta number) is a dimensionless number:

-33 2 2/

2

OO HO HO

L

k C DHa

k

−

= Eq. 17

Di is the liquid diffusion coefficient of any species i. If the reaction kinetics is fast in the liquid film (Ha

>3), tanh(Ha) tends towards 1. Therefore:

-33 2

3 3

,

2 ,/

,

ln [ ]

G

A i

L O LG O HO

A o O G O G

C VRT VRTHak a k HO D a

C H Q H Q

−= ° = ° Eq. 18

Consequently:

( )3 3

-33 2

2/

ln 1

[ ]

O G O

OO HO

Eff Q Ha

VRTk HO D−

− −° =

Eq. 19

In a given gas-liquid mechanically agitated vessel, a° depends on the gas flow rate, impeller speed

and liquid properties (ionic strength, presence of surfactant). Consequently, these parameters must

be controlled. Since a° does not depends on the absorbed species, extrapolation from O3 to H2S

chemical system is not an issue. However, the ionic strength must be kept low to have identical

properties between both chemical systems. a° has been determined for various gas flow rates and

H2O2 concentrations at pH = 10. The HO2- concentration in the contactor was deduced from the pH

measured and the H2O2 concentrations at the inlet, and the mass balance considering that one mol

of O3 transferred consumed one mol of H2O2.

In the thin range of studied superficial velocity, a° is not statistically dependent on the velocity (Table

A.3 in appendix). An average value of 76.1 ± 4.9 m2 m

-3 is found (Relative Standard Deviation RSD =

6%). The average value of a° is in agreement with the order of magnitude expected. The standard

deviation stays within the calculated 10% experimental uncertainty. The parity plot (Fig. 2) confirms a

good agreement between the experimental and theoretical (using Eq. 19 and a° = 76.1 m2 m

-3) EffO3

values. The values of the Hatta number, using the kL value measured later (§ 4.2), are larger than 3,

confirming the fast reaction assumption (Table A.3 in appendix).

Figure 2: Parity diagram comparing the experimental values of the ozone removal efficiency with the theoretical values

obtained with a° = 76.1 m2 m

-3 for various H2O2 concentrations.

4.2 Liquid-side mass transfer coefficient determination

The liquid side mass transfer coefficient kL depends on the absorbed species diffusion coefficient, the

liquid properties (viscosity, density) and the liquid hydrodynamics. kL can be determined by chemical

or physical methods (Hoffmann et al., 2007; Rejl et al., 2009). Chemical methods results can be

biased by the reactant added which can affect the liquid properties. The only restriction linked to the

physical absorption method is the requirement of a poorly soluble compound to avoid reaching the

gas-liquid equilibrium and limit the gas-phase resistance (RG →0) (Lara Marquez et al., 1994).

Consequently, the physical absorption of pure oxygen at steady-state has been selected. In this case:

2

2 2

2

, ,

OEq Eq

O i O o

O

pC C

H= =

Eq. 20

Therefore, since the liquid phase is perfectly mixed, the driving force is constant in the whole volume

of the gas-liquid contactor:

( )− = −2

2 2 2

2

,

OEq

O O O o

O

pC C C

H Eq. 21

The mass transfer equation (Eq. 8) coupled with the mass balance equation leads to Eq. 22:

15%

20%

25%

30%

35%

40%

15% 20% 25% 30% 35% 40%

Eff t

he

o

Effexp

1173 mg/L

1470 mg/L

1683 mg/L

1794 mg/L

2720 mg/L -10%

+10%

( ) = ° − = −

2

2 2 2

2

, , ,

O

L O o L O o O i

O

pN k a V C Q C C

H Eq. 22

( )−⇒ =

° −

2 2

2

2

2

, ,

,

L O o O i

L

O

O o

O

Q C Ck

pa V C

H

Eq. 23

QL is the liquid volume flow rate (m3 s

-1).

2

A

O

p

H corresponds to the O2 concentration at saturation in

the liquid which can be measured in the semi-continuous mode waiting for the saturation. No

significant effect of the gas and liquid flow rates on kL has been emphasized (Table 2). An average

value of 2O

Lk = 2.7×10

-4 ± 0.25×10

-4 m s

-1 (RSD = 9%) was found which is totally in agreement with the

expected value. The H2S liquid side mass transfer coefficient must be extrapolated from the oxygen

one by the following relation (Roustan, 2003):

2

2 2

2

2

4 4 12.5 10 0.2 10 m s

H SH S H SL

LO

L O

Dkk

k D

− − −= ⇒ = × ± × Eq. 24

Table 2: Results of the oxygen absorption experiments.

QG (NL h-1

) 426 426 469 469 469 509 509 509

USG (cm s-1

) 1.24 1.24 1.37 1.37 1.37 1.49 1.49 1.49

QL (L h-1

) 21.2 28.6 13.9 21.2 28.6 13.9 21.2 28.6

104×kL (m s

-1) 2.29 2.55 2.43 2.89 2.80 2.89 2.79 2.98

4.3 Gas-side mass transfer coefficient determination

To measure the gas side coefficient, the liquid resistance must be negligible (RL →0) which is the case

when E is very large due to a very fast chemical reaction with a large excess of the reagents.

Moreover, the solute is instantaneously consumed at the gas-liquid interface (CA = 0). According to

Eq. 11 :

( ),

,

ln ln 1

G

A i G G

GG

A o G

C k a V Qk Eff

C Q a V

°= ⇒ = − −°

Eq. 25

kG has been measured by H2S absorption in an alkaline chlorine solution (10.5 < pH < 11 and CClOH = 1

g L-1

) for various gas and liquid flow rates (500 < QG < 550 NL h-1

). No significant effect of the flow

rates on kG has been emphasized. An average value of kG = 4.2×10-3

± 0.33×10-3

m s-1

(RSD = 8%) was

found.

The kG value is a rather low and comparable to the values found in packed or spraying columns. Since

it has never been measured for a mechanically agitated vessel, there is no possibility to discuss the

validity of this result. In the case of H2S, which is rather poorly soluble in water, the mass transfer

resistance in the gas side (RG) must not be neglected, especially with increasing E values in the liquid

phase (Fig. 3). This result is very important for the following step. Indeed, the enhancement factor

obtained during the H2S absorption in chlorine solution should not be too large to limit as much as

possible the gas-side resistance and improve the sensitivity of the calculations.

Figure 3: Liquid-phase relative mass transfer resistance RL and overall liquid-phase mass transfer coefficient KL vs. the

enhancement factor E.

4.4 Validation of the mass transfer and hydrodynamic parameters

4.4.1 H2S absorption in slightly alkaline water

The accuracy of a°, kL and kG values has been evaluated for H2S absorption in water at alkaline pH

(NaOH was added) using the operating conditions synthesized in Table 1. The pH was slightly alkaline

to improve the H2S removal efficiency (>20%) at a level sufficient to limit the experimental

uncertainty.

Without H2S oxidation and assuming that S2-

is negligible at the working pH, CA can be deduced from

the mass balance equation:

0

2

4

6

8

10

12

14

16

18

20

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 10 100 1 000

10

4×

KL(m

s-1

)

RL

E

( ) ( ),2

2 2 2, , 1 10 ApH pKG G

G H S i H S o L H SN Q C C Q C

−= − = + Eq. 26

According to Eqs 11 and 26:

( )( )( )

−

−

+ − = ° + − −

,2

2

,2

2 2

1 10ln

1 10 1

A

A

pH pK

H S

LpH pK

LH S H S

A Eff AVK a

QA Eff Eff Eq. 27

A is the absorption rate (no unit) defined by Eq 28:

2

L

G H S

Q RTA

Q H=

Eq. 28

This equation provides a limit value of EffH2S since the term in the logarithm must be positive:

( )( )

,2

,2max

1 10

1 1 10

A

A

pH pK

pH pK

AEff

A

−

−

+=

+ + Eq. 29

Effmax is obtained when the chemical equilibrium is reached between the gas at the outlet and the

liquid. It is the maximal removal efficiency for given pH, temperature and gas and liquid flow rates.

Finally, from Eq. 29:

( )

( )( )

−

−

+ − °

=

− + + °

,2

2

,2

1 10 1 exp

1 1 10 1 exp

A

A

pH pK

L

L

H S

pH pK

L

L

AVA K a

QEff

AVA K a

Q

Eq. 30

H2S mass transfer in water can be enhanced thanks to the H2S dissociation (Ediss) and/or

recombination with HO- (EHO-) which form HS

- in the liquid film (Biard and Couvert, 2013). This

phenomenon is highly influenced by the pH. Assuming that these reactions are instantaneous

comparing to mass transfer, the enhancement factors for each reaction can be calculated for each

experimental point (Biard and Couvert, 2013). The global enhancement factor can be calculated for 2

instantaneous reactions by the sum of the individual enhancement factor minus 1 (Chang and

Rochelle, 1982; van Swaaij and Versteeg, 1992). Then, it leads to:

( ) ( )( ) ( )( )

-

-

-2

2 2w A,2

2 2

2 2 2

dissHO

HOH S *HO

H SpK -pK

2*

diss *

H S

1

101

10

10 4 1 10 101

2

wpH pK

HO

H S

HS

pH pH pH

H S H SH HS HS H HS HS H HS HS

H S H S

E E E

DE

D DD C

D

D D C C C D D C D D CE

D C C

−

−

+ − − + − − + − −

∞

−

− − −

= + −

= ++

×

+ − − − += +

−

Eq. 31

pkW is the negative log of the water ion product. The dissociation enhancement factor is low and

included in the range 1.01-1.05 while the recombination with HO- enhancement factor is in the range

1.18-1.29.

Comparison between the experimental removal efficiency and the theoretical values deduced from

Eqs 30 and 31 is really good (Fig. 4). For low values of A, the equilibrium is almost reached since the

removal efficiency is rather close of the maximal removal efficiency obtained from Eq. 29.

Figure 4: (a). H2S removal efficiency vs. A(1+10pH-pKA,2) for various gas flow rates without chlorine. The colored lines

represent the prediction of the model (Eq. 30) and the black line represents the prediction of the maximal removal

efficiency (Eq. 29). (b). Parity plot between the experimental and predicted removal efficiencies.

4.4.2 Sensitivity analysis

The good agreement between the experimental and the theoretical values is not a sufficient

condition to confirm the accuracy of the mass transfer coefficients and interfacial area determined

previously and the robustness of the model. Indeed, each parameter has a variable influence on the

removal efficiency. This influence can be assessed through a sensitivity analysis based on the

elasticity index determination. The sensitivity analysis quantifies the effect of possible change in

inputs on the model output which is here the removal efficiency (Asadi Kalameh et al., 2012; Yin et

al., 2005). Consequently, it addresses the relative significance of potential errors in various input

parameters. The elasticity index EIP/Eff measures the relative change of the removal efficiency for a

relative change in an input parameter P. For example, for the input kL :

L L L L

L

for k for k for k for k

/

for k

1

2

L L

L

k kL

k Eff

L L

Eff Eff Eff EffkEI

Eff k k

+∆ −∆− − = × + ∆ −∆ Eq. 32

10%

15%

20%

25%

30%

35%

40%

45%

50%

10% 15% 20% 25% 30% 35% 40% 45% 50%

Eff

exp

Efftheo

-10%

+10%

20%

25%

30%

35%

40%

45%

50%

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2

Eff

exp

A(1+10pH-pKA,2)

1.32 cm/s

1.44 cm/s

1.55 cm/s

Eff max

USG = 1.32 cm s-1

USG = 1.44 cm s-1

USG = 1.55 cm s-1

EffMax

(a) (b)

EffH2S must be a monotonic function of the input parameter P. The average elasticity indexes, taking

into account all the experimental points, have been estimated for all the input of Eq. 30 for a relative

change of 10% (Table 3). The indexes for the input QG and HH2S are negative since Eff is a decreasing

function of them. Except for the pH, the standard deviation is low, demonstrating that EI is almost

constant. Results of Table 3 emphasize that Eff is mainly sensitive to two measured variables, pH and

QG, and two physico-chemical constants, pKA,2 and HH2S, whose the values taken from the literature

are consistent between many authors and should be precise. Consequently, the difference between

the experimental and the theoretical values of EffH2S (Fig. 4) might be mainly due to the experimental

errors of the pH, and in a lower extent of QG, more than uncertainties of kL, kG and a° determined

previously. On the one hand, pH was rather unsteady and affected by the gas bubbles with a low

precision of ± 0.1 unit. On the other hand, QG was measured with a limited combined uncertainty of

3.1% taking into account the uncertainties of the variable area flowmeter, the pressure and the

temperature measurements used for the flow rate correction. The strong sensitivity of Eff to the pH

and pkA,2 can be explained by the fact that these inputs affect significantly the difference of H2S

concentrations between the interface and the bulk (mass transfer driving force).

Table 3: Elasticity index determination for EffH2S and KLa°.

Removal efficiency (EffH2S)

Input parameter pH pKA,2 QG HH2S a° and V E and kL QL kG

Average EI 4.83 2.30 -0.81 -0.73 0.45 0.37 0.36 0.079

Standard deviation 1.42 0.93 0.04 0.05 0.07 0.06 0.11 0.013

Overall liquid mass transfer coefficient (KLa°)

Input parameter E and kL kG

Average EI 0.83 0.17

Standard deviation 0.004 0.004

The removal efficiency (EffH2S) is also significantly influenced by a° and V (which possesses a low

experimental uncertainty of 3.0%), proving the a° value determined previously is coherent and

confirming the order of magnitude. The EI determination for E and kL shows that EffH2S is moderately

sensitive to these parameters. Consequently, the order of magnitude of E and kL should be coherent.

On the contrary, EffH2S is almost independent of kG since the gas phase resistance is low (15% for E

close to 1 according to Fig. 3). This is confirmed by the sensitivity index calculation of KLa° for the

input kG (EIkLa/kG) which is 5 times lower than EIKLa/kL

and EIKLa/E (Table 3). By the way, the kG value

found previously cannot be validated. To limit the influence of this uncertainty on the E

determination using chlorine, the gas-phase resistance should be kept low as demonstrated by Fig. 3.

Moreover, Fig. 3 highlights an asymptotic behavior which emphasizes that KLa° becomes less

sensitive to E when E increases which could affect the uncertainty of the kinetic study.

4.5 Oxidation kinetics investigation

4.5.1 Enhancement factor determination

Table 4: Results of the H2S reactive absorption in chlorine solutions.

pH CNaOCl

(ppm) 2

G

H SC (10 3 mol

m-3

)

y†

(%)

Eexp Etheo E39/E39+40‡

(%)

Effexp

(%)

Efftheo

(%)

Error

(%)

4.90 293 1.50 84.3 7.0 5.41 38.5 89.0 85.6 4.0

4.92 301 1.65 83.2 6.4 5.49 38.8 88.0 86.1 2.3

4.93 305 1.43 85.3 7.5 5.55 39.5 89.7 86.0 4.3

4.94 222 2.50 73.9 3.6 4.69 38.6 79.3 84.3 5.9

4.95 229 2.38 75.2 3.8 4.78 39.1 80.4 84.4 4.8

4.95 233 2.01 78.5 4.8 4.83 39.7 83.6 83.8 0.2

5.59 278 1.19 88.6 4.2 6.64 67.4 79.4 86.8 8.5

5.59 282 1.25 88.1 3.9 6.66 67.0 78.1 86.8 10.0

5.58 284 1.28 87.9 3.8 6.66 66.7 77.8 86.9 10.5

5.80 45 2.17 70.0 2.3 2.86 74.8 68.7 74.4 7.8

5.88 170 2.05 80.6 5.2 6.01 76.9 85.1 87.2 2.4

5.88 178 2.12 80.4 5.0 6.15 76.8 84.3 87.5 3.5

5.88 182 2.07 80.9 5.2 6.23 76.9 85.2 87.8 3.0

5.99 351 1.78 87.8 9.6 9.64 82.1 90.0 90.1 0.0

5.99 365 1.72 88.3 10.0 9.85 82.2 90.3 90.2 0.1

6.04 202 1.21 90.1 9.6 7.76 84.5 92.0 90.1 2.1

6.50 187 1.47 91.2 6.9 11.25 93.8 88.7 93.1 4.7

6.51 191 1.45 91.5 7.0 11.48 94.0 88.9 93.2 4.6

6.94 269 0.39 98.6 27.8 21.13 97.9 96.6 95.9 0.7

7.53 257 0.46 98.9 23.6 32.10 99.4 96.2 96.8 0.7

7.53 265 0.25 99.4 44.1 32.67 99.4 97.3 96.8 0.4

Chemical absorption of H2S was realized at various chlorine concentration and pH (Table 4). Since the

H2S oxidation rate is higher than the H2S mass transfer rate and the Damkölher number is large (10 to

30), CA can be neglected comparing toEq

AC (Roustan). In this case, according to Eq. 11, KLa° and Eexp

can be deduced for each experiment by respectively Eqs 33 and 7:

( )2 2

2

2 2

,

,

ln ln 1

G

H S i H S G

L L H SG

H S o H S G

C H QVRTK a K a Eff

C H Q VRT= ° ⇒ ° = − −

Eq. 33

† x was comprised between 0.98 and 1.00 demonstrating that the reaction with ClOH is pseudo-first order.

‡ Represents the ratio of E obtained considering only the reaction 39 on E considering both reactions 39 and 40.

The results show that Eexp (i.e. the oxidation kinetic rate) increases obviously with the chlorine

concentration but also surprisingly with the pH. Indeed, ClOH is considered as more reactive than

ClO- which lets expect a larger kinetic rate for lower pH. It might emphasize that HS

- plays a role in

the oxidation mechanism, even if at the implemented pH, the HS- concentration is significantly lower

than the H2S concentration. We note that Eexp values were included in the range 2.3-44, confirming

that H2S is entirely consumed in the liquid film for most experiments.

4.5.2 Kinetic constants determination

According to the experimental results, the oxidation mechanism might involve both acid-base and

oxidation reactions. Indeed, H2S and HS- can be both oxidized by ClO

- or ClOH. A list of the potential

reactions is presented in Table 5.

According to Eigen (1964), k1 and k3 values should be close to 3×1010

and 109 L mol

-1 s

-1. The

increasing of the kinetic rate with the pH can be due to the fact that a part of H2S reacts with ClO- in

the liquid film to form HS- (reaction 36) which can be further oxidized by ClOH (oxidation with ClO

- is

unexpected according to the order of magnitude of k4 and k5). Moreover, ClOH can oxidize directly

H2S. If we assume that reverse reaction 36 is negligible, then the whole mechanism can be reduced

to the two irreversible reactions 39 (equivalent to reaction 36 followed by reaction 37) and 40; and

the two kinetic constants k6 and k7 must be determined.

Table 5 : H2S oxidation mechanism. k-2 value and k1 and k3 orders of magnitude were obtained from (Eigen, 1964). k4 and

k5 orders of magnitude were obtained from (Deborde and von Gunten, 2008).

Reaction Type Equilibrium constant

Eq. 34 Acid-base A,2 w-pK +pK1

1

10k

Kk−

= =

Eq. 35 Acid-base A,2-pK2

2

10k

Kk−

= =

Eq. 36 Acid-base A,1 A,2pK - pK3

3

10k

Kk−

= =

Eq. 37 HS

- oxidation

Eq. 38 HS

- oxidation

Eq. 39 H2S oxidation

Eq. 40 H2S oxidation

The enhancement factor for two irreversible parallel reactions in the liquid film can be determined

using the Onda and coworkers approximated solution for the two-film theory (Biard et al., 2010;

Onda et al., 1970). From the material balances and the boundary conditions:

− − − − + − + − =

−

2

2 2 2 2

2 2

**

* *

*

1 1

1

H S ClOH ClOH ClOH ClO ClO ClO

H S H S H S ClOH H S ClOH ClOH

H S H S

C D C CC D C

C C D C D C CE

C C Eq. 41

Due to the fast oxidation of H2S in solution, the bulk concentration is negligible compared to the

interface concentration:

( ) ( )

2 2 2

2 2 2 2

**

*

**

* *

1 1

1 1 1 with = and =

ClOH ClOH ClOH ClO ClO ClO

H S H S ClOH H S ClOH ClOH

ClOH ClOH ClO ClO ClOH ClO

H S H S H S H S ClOH ClO

D C CC D CE

C D C D C C

D C CD C CE x y x y

D C D C C C

− − −

− − −

−

= + − + −

⇒ = + − + − Eq. 42

Onda et al. (1970) proposed an approximate solution of the differential equations system obtained

by linearization. For two second-order reactions:

( )( ) ( ) ( )

− = ≈

−

2

2

2 2

2 *

22

*

2 2

11

cosh

1 tanh tanh

H S

H S

H S H S

CM

C M ME

C C M M Eq. 43

With:

− = +

27 6

2 2

7

H S ClOH ClO

L ClOH

k D C CkM x y

k k C Eq. 44

Eqs. 42 and 43 are two independent equations with 3 variables (x, y and E). Therefore, an additional

equation is required. Onda et al. (1970) assumed that the reagent concentrations are quadratic

functions of the film thickness and obtained the following equation:

( )−

−

− −=

−+

6

7

6

7

11 1

6

5 11

6

ClOH

ClO

ClOH

ClO

D kx

D ky

D k x

D k x

Eq. 45

Following the numerical procedure developed in the supplementary material, k6 (6.75 106 L mol

-1 s

-1)

and k7 (1.62 105 L mol

-1 s

-1) were determined by the least square method trying to minimize the error

between the experimental and theoretical values of E (procedure presented in the supplementary

material). These values confirm the order of magnitude expected by Biard et al. (Biard et al., 2010).

The determined apparent kinetic constant of H2S with ClO- is larger than the one with ClOH which

corroborates the assumption that H2S can react with ClO- through an acid base reaction before

oxidation. To assess the importance of both reactions with ClO- and ClOH, the ratio (E39/E39+40) of the

enhancement factor calculated assuming just the reaction with ClO- on the overall enhancement

factor has been calculated (Table 4). It confirms that, except for circumneutral and basic pH, both

reactions contribute to the overall mass transfer enhancement.

The relative errors between the experimental and theoretical values of EffH2S are summarized in

Table 4 and confirm a rather good agreement of the model. In the one hand, the discrepancy

between the experimental and theoretical values of E is rather large (average error of 24%) since the

determination of E is very sensitive to the experimental uncertainties (on Eff, QG and the pH

especially). In other words, even low experimental uncertainties induce a rather large uncertainty on

E. However, the points are distributed on both sides of the parity plot excluding systematic errors

(Fig. 5). On the other hand, calculations of theoretical values of EffH2S are poorly sensitive to E. By the

way, the difference between the experimental and theoretical values of EffH2S is really low (3.8% in

average). These observations are emphasized by the sensitivity analysis (§ 4.5.3).

Figure 5: Parity plot between the experimental and predicted removal efficiencies during H2S absorption in chlorine

solutions.

At basic pH, reaction 34 should not be neglected anymore. A competition between Eq. 34 and 36 is

expected depending on the pH and the total chlorine concentration. In the next step of this project,

k6 and k7 will be confronted to the kinetic constants that we measured for HS- oxidation by chlorine

0.6

0.7

0.8

0.9

1.0

0.6 0.7 0.8 0.9 1.0

Eff

the

o

Effexp

+10%

-10%

(result not published yet) to build a kinetic model at basic pH. More complex numerical resolution

tools will be required to solve the high numbers of differential equations and finally build a model

which can provide an efficient tool for scrubbers design.

4.5.3 Sensitivity analysis

Figure 6: The EffH2S Elasticity Index vs. the enhancement factor for several inputs.

The sensitivity analysis quantifies the influence of the input E, kL, kG, V and T on the output EffH2S (Fig.

6) for increasing values of E. The three most significant inputs are E, kL and kG which directly influence

the mass transfer rate. Errors on the enhancement factor determination (due to uncertainties on k6

and k7) are more significant for low values of E with more liquid phase resistance. With significant

chlorine concentrations up to 500 ppm at alkaline pH, the kinetics of oxidation is so large that the

resistance in the liquid phase is almost negligible and the mass transfer rate is controlled by the gas

phase. This conclusion obtained with a well agitated reactor can be extrapolated to packed column

widely used in gas treatment. Therefore, designers must adjust their concentration to insure an

excess of chlorine in order to maintain a low liquid phase resistance in the whole gas-liquid

contactor, especially at the bottom where chlorine has been partly consumed.

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0 10 20 30

Ela

stic

ity

Ind

ex

E

E, kL and kG

V, T, H and QG

5 Conclusions

Kinetics of H2S reaction with hypochlorite anion and hypochlorous acid was investigated by reactive

absorption in a well agitated gas-liquid contactor. The liquid and gas sides mass transfer coefficients

as well as the interfacial area has been determined using appropriated experimental conditions.

These values were confirmed with the modeling of reactive absorption of H2S in water without

oxidant. Absorption of H2S in chlorine solutions demonstrated that the oxidation kinetic rate

increases surprisingly with the pH which highlights that HS- may play a role in the oxidation

mechanism. We assumed that H2S reacts with ClO- to form HS

- which is further oxidized by ClOH to

build a global mechanism based on the H2S irreversible reactions with ClO- (k6 = 6.75 10

6 L mol

-1 s

-1)

and ClOH (k7 = 1.62 105 L mol

-1 s

-1). A good agreement between the experimental and the theoretical

values of the removal efficiencies confirmed the order of magnitude of the kinetic constants found

and the validity of the model at the pH range tested. The next step of this study will require to

improve the model for the description of the mechanism at basic pH.

6 Acknowledgements

The authors wish to thank Candelà De La Sota Sandez for her participation to the RTD measurements

and Dominique Allaire for his contribution to the set-up construction. We also wish to thanks

Christophe Renner from Veolia Environnement Recherche et Innovation (Maisons-Laffitte) for his

material contribution to this project.

References

Aieta, E.M., Roberts, P.V., 1986. Application of mass transfer theory to the kinetics of a fast

gas-liquid reaction: chlorine hydrolysis. Environ. Sci. Technol. 20, 44-50.

Asadi Kalameh, H., Karamali, A., Anitescu, C., Rabczuk, T., 2012. High velocity impact of metal

sphere on thin metallic plate using smooth particle hydrodynamics (SPH) method. Front. Struct. Civ.

Eng. 6, 101-110.

Beltrán, F.J., 2004. Ozone reaction kinetics for water and wastewater systems. CRC Press, Boca

Raton.

Biard, P.-F., 2009. Contribution au développement d'un procédé de lavage chimique compact.

Traitement du sulfure d'hydrogène par le chlore à l'échelle semi-industrielle et de COV odorants par

oxydation avancée O3/H2O2 à l'échelle du laboratoire., ENSCR. University of Rennes 1, Rennes, p. 229.

Biard, P.-F., Couvert, A., 2013. Overview of mass transfer enhancement factor determination

for acidic and basic compounds absorption in water. Chem. Eng. J. 222, 444-453.

Biard, P.-F., Couvert, A., Renner, C., Levasseur, J.-P., 2010. Wet scrubbing intensification

applied to hydrogen sulphide removal in waste water treatment plant. Can. J. Chem. Eng. 88, 682-

687.

Biard, P.-F., Couvert, A., Renner, C., Zozor, P., Bassivière, S., Levasseur, J.-P., 2009. Hydrogen

sulphide removal in waste water treatment plant by compact oxidative scrubbing in Aquilair PlusTM

process. Wat. Practice. technol. 4, doi:10.2166/wpt.2009.2023.

Bonnin, C., 1991. Les sources de nuisances olfactives dans les stations de traitement des eaux

usées résiduaires, et leur traitement par lavage à l'eau chlorée en milieu basique, Ecole Nationale

Supérieure de Chimie de Rennes. Université de Rennes I, Rennes, p. 191.

Brettschneider, O., Thiele, R., Faber, R., Thielert, H., Wozny, G., 2004. Experimental

investigation and simulation of the chemical absorption in a packed column for the system NH3–CO2–

H2S–NaOH–H2O. Sep. Purif. Technol. 39, 139-159.

Busca, G., Chiara, P., 2003. Technologies for the abatement of sulphide compounds from

gaseous streams: a comparative overview. J. Loss. Prev. Process Indust. 16, 363-371.

Calderbank, P., 1958. Physical rate processes in industrial fermentation - Part I: The interfacial

area in gas-liquid contacting with mechanical agitation. Chem. Eng. Res. Des. 36, 443-463.

Calderbank, P.H., 1959. Physical rate processes in industrial fermentation - Part II: Mass

transfer coefficients in gas-liquid contacting with and without mechanical agitation. Chem. Eng. Res.

Des. 37, 173-185.

Chang, C.S., Rochelle, G.T., 1982. Mass transfer enhanced by equilibrium reactions. Ind. Eng.

Chem. Fund. 21, 379-385.

Chao, M., 1968. The diffusion coefficients of hypochlorite, hypochlorous acid, and chlorine in

aqueous media by chronopotentiometry. J. Electrochem. Soc. 115, 1172.

Chen, L., Huang, J., Yang, C.-L., 2001. Absorption of H2S in NaOCl caustic aqueous solution.

Environ. Prog. 20, 175-181.

Choppin, A.R., Faulkenberry, L.C., 1937. The oxidation of aqueous sulfide solutions by

hypochlorites. Journal of the American Chemical Society 59, 2203-2207.

Deborde, M., Von Gunten, U., 2008. Reactions of chlorine with inorganic and organic

compounds during water treatment - Kinetics and mechanisms: A critical review. Wat. Res. 42, 13-51.

Ebrahimi, S., Picioreanu, C., Kleerebezem, R., Heijnen, J.J., van Loosdrecht, M.C.M., 2003. Rate-

based modelling of SO2 absorption into aqueous NaHCO3/Na2CO3 solutions accompanied by the

desorption of CO2. Chem. Eng. Sci. 58, 3589-3600.

Eigen, M., 1964. Proton transfer, acid-base catalysis, and enzymatic hydrolysis. Part I:

elementary processes. Angewandte Chem. Int. Edition in English 3, 1-19.

Fernández-Prini, R., Alvarez, J.L., Harvey, A.H., 2003. Henry's constants and vapor-liquid

distribution constants for gaseous solutes in H2O and D2O at high temperatures. J. Phys. Chem. Ref.

Data 32, 903-916.

Foussard, J.-N., Debellefontaine, H., 2000. Thermodynamic basis for the solubility and

diffusivity of ozone in water, IOA International specialised symposium, fundamental and engineering

concepts for ozone reactor design, Toulouse, pp. 35-38.

Gostelow, P., Parsons, S.A., Stuetz, R.M., 2001. Odour measurements for sewage treatment

works. Wat. Res. 35, 579-597.

Hikita, H., Asai, S., 1964. Gas absorption with (m, n)th

order irreversible chemical reaction. Int.

Chem. Eng. 4, 332–340.

Hoffmann, A., Mackowiak, J., Gorak, A., Haas, M., Loning, J., Runowski, T., Hallenberger, K.,

2007. Standardization of mass transfer measurements: a basis for the description of absorption

processes. Chem. Eng. Res. Des. 85, 40.

Jia, Y., Zhong, Q., Fan, X., Wang, X., 2010. Kinetics of oxidation of total sulfite in the ammonia-

based wet flue gas desulfurization process. Chem. Eng. J. 164, 132-138.

Jing, G., Zhou, L., Zhou, Z., 2012. Characterization and kinetics of carbon dioxide absorption

into aqueous tetramethylammonium glycinate solution. Chem. Eng. J. 181–182, 85-92.

Joshi, J., Pandit, A., Sharma, M., 1982. Mechanically agitated gas-liquid reactors. Chem. Eng.

Sci. 37, 813-844.

Kangas, J., Jäppinen, P., Savolainen, H., 1984. Exposure to hydrogen sulfide, mercaptans and

sulfur dioxide in pulp industry. AIHA J. 45, 787-790.

Kerc, A., Olmez, S.S., 2010. Ozonation of odorous air in wastewater treatment plants. Ozone

Sci. Eng. 32, 199 - 203.

Kohl, A.L., Nielsen, R., 1997. Gas purification. Gulf Professional Publishing.

Kucka, L., Kenig, E.Y., Górak, A., 2002. Kinetics of the gas-liquid reaction between carbon

dioxide and hydroxide ions. Ind. Eng. Chem. Res. 41, 5952-5957.

Langlais, B., Reckhow, D.A., Brink, D.R., 1991. Ozone in water treatment. Chelsea, MI (United

States) ; Lewis Publishers, United States.

Lara Marquez, A., Wild, G., Midoux, N., 1994. A review of recent chemical techniques for the

determination of the volumetric mass transfer coefficient kLa in gas-liquid reactor. Chem. Eng. Proc.

33, 247-260.

Le Sauze, N., Laplanche, A., Martin, G., Paillard, H., 1991. A process of washing and ozonation

to deodorize an atmosphere contaminated by sulfides. Ozone Sci. Eng. 13, 331-347.

Lee, S.Y., Pang Tsui, Y., 1999. Succeed at gas/liquid contacting. Chem. Eng. Prog. 95, 23-49.

Morris, J.C., 1967. Kinetics of reactions between aqueous chlorine and nitrogen compounds.

John Wiley & Sons, New York, NY.

Onda, K., Sada, E., Kobayashi, T., Fujine, M., 1970. Gas absorption accompanied by complex

chemical reactions—III Parallel chemical reactions. Chem. Eng. Sci. 25, 1023-1031.

Perry, R.H., Green, D.W., 1997. Perry's chemical engineers' handbook, 7th

edition. McGraw-Hill,

New-York.

Rappert, S., Müller, R., 2005. Odor compounds in waste gas emissions from agricultural

operations and food industries. Waste Manag. 25, 887-907.

Rejl, J., Linek, V., Moucha, T., Valenz, L., 2009. Methods standardization in the measurement of

mass transfer characteristics in packed absorption columns. Chem. Eng. Res. Des. 87, 695-704.

RESI, 2013. Gas-liquid mass transfer in gassed mechanically agitated vessels.

http://www.resi.com.tw/Mixing/08DOC.pdf.

Roustan, M., 2003. Transferts gaz-liquide dans les procédés de traitement des eaux et des

effluents gazeux. Lavoisier, Paris.

Sema, T., Naami, A., Fu, K., Edali, M., Liu, H., Shi, H., Liang, Z., Idem, R., Tontiwachwuthikul, P.,

2012. Comprehensive mass transfer and reaction kinetics studies of CO2 absorption into aqueous

solutions of blended MDEA–MEA. Chem. Eng. J. 209, 501-512.

Smet, E., Langenhove, H.V., 1998. Abatement of volatile organic sulfur compounds in odorous

emissions from the bio-industry. Biodegradation 9, 273-284.

Smet, E., Lens, P., Langenhove, H.V., 1998. Treatment of waste gases contaminated with

odorous sulfur compounds. Crit. Rev. Environ. Sci. Tech. 28, 89-117.

Sotelo, J.L., Beltrán, F.J., Gonzalez, M., Garcia-Araya, J.F., 1991. Ozonation of aqueous solutions

of resorcinol and phloroglucinol. 2. Kinetic study. Ind. Eng. Chem. Res. 30, 222-227.

Staehelin, J., Hoigne, J., 1982. Decomposition of ozone in water: rate of initiation by hydroxide

ions and hydrogen peroxide. Environ. Sci. Technol. 16, 676-681.

Tamimi, A., Rinker, E.B., Sandall, O.C., 1994. Diffusion coefficients for hydrogen sulfide, carbon

dioxide, and nitrous oxide in water over the temperature range 293-368 K. J. Chem. Eng. Data 39,

330-332.

Vaidya, P.D., Kenig, E.Y., 2007a. Absorption of CO2 into aqueous blends of alkanolamines

prepared from renewable resources. Chem. Eng. Sci. 62, 7344-7350.

Vaidya, P.D., Kenig, E.Y., 2007b. Gas–Liquid reaction kinetics: a review of determination

methods. Chem. Eng. Comm. 194, 1543-1565.

Vaidya, P.D., Kenig, E.Y., 2009. A Study on CO2 absorption kinetics by aqueous solutions of N,

N-Diethylethanolamine and N-Ethylethanolamine. Chem. Eng. Technol. 32, 556-563.

van Swaaij, W.P.M., Versteeg, G.F., 1992. Mass transfer accompanied with complex reversible

chemical reactions in gas—liquid systems: an overview. Chem. Eng. Sci. 47, 3181-3195.

Wang, T.X., Margerum, D.W., 1994. Kinetics of reversible chlorine hydrolysis: Temperature

dependence and general-acid/base-assisted mechanisms. Inorg. Chem. 33, 1050-1055.

Whitman, W.G., 1923. A preliminary experimental confirmation of the two-film theory of gas

absorption. Chem. Met. Eng. 29, 146-148.

Yin, H., Wu, S., Ejeta, M., 2005. CalSim-II Model Sensitivity Analysis Study: Technical

Memorandum Report. california Department of Water Resources, Bay-Delta Office.

Appendix A: Additional tables

Table A. 1 : Compilation of the physico-chemical properties used.

Parameter Values at 293 K References

Water

ρL (kg m-3

) 1000

ML (kg mol-1

) 0.018

pKw 14.15

H2S

HH2S (Pa m3

mol-1

) 864

(Fernández-Prini et al.,

2003)

DH2S (m2 s

-1) 1.75×10

-9 (Tamimi et al., 1994)

pKA,2 (H2S/HS-

)

3131log 0.00361exp

T

− −

=> 7.07 at 293 K (Roustan, 2003)

pKA,3 (HS-/S

2-)

7211.2log exp 7.489

T

− − −

=> 13.94 at 293 K (Brettschneider et al.,

2004)

O3

HO3 (Pa m

3

mol-1

)

5 3 -14030exp 22.3 1.013 10 9400 Pa.m .molL

L

M

T ρ − × × × =

(Langlais et al., 1991)

DO3 (m

2 s

-1) 1.7×10

-9

(Foussard and

Debellefontaine, 2000)

kO3/HO2- (m

3

mol-1

s-1

) 2.8×10

9 ± 0.5×10

9

(Staehelin and Hoigne,

1982)

O2

HO2 (Pa m

3

mol-1

) 7.3×10

4 (Roustan, 2003)

DO2 (m

2 s

-1) 2.1×10

-9 (Roustan, 2003)

ClOH/ClO-

K1 (Cl2 →

HClO) ( )

0.5 0.5

0.5 0.5

53 33000 0.5exp exp 1.02 0.23 0.5

1 1 0.5

II

R RT I

− + − + − + +

=> 3.8×10

-4 at I = 0 et 293 K

Eq. deduced from (Wang

and Margerum, 1994)

pKA,1

(ClOH/ClO-)=

300010.0686 0.0253 T

T− + × => 7.58 at 293K (Morris, 1967)

DClO- (m2 s

-1)

10605.40

10 T

− −=> 0.96×10

-9 at 293K (Chao, 1968)

DClOH (m2 s

-1)

9455.72

10 T

− −=> 1.13×10

-9 at 293K (Chao, 1968)

H2O2/HO2-

pKA,4 11.8 (Beltrán, 2004)

Table A.2: Suppliers and references of the chemical products and equipments used.

Chemical products

NaOH 10 mol L-1

Labogros France

H3PO4 85% wt Acros Organics Belgium

NaOCl 13% wt Acros Organics Belgium

H2O2 35% wt Acros Organics Belgium

Acetic acid 99.9% Acros Organics Belgium

Indigo carmine Analytical grade Acros Organics Belgium

Sodium thiosulfate Titrisol® 0.1 mol L-1

Merck Germany

Potassium iodide Analytical grade Acros Organics Belgium

Ammonium molybdate 98.5 0025 Rectapur VWR Belgium

O2 (cylinder) Pure Air Liquide France

N2/H2S mixture

(cylinder) 1000 ± 50 ppmv in H2S Air liquide France

Equipments

Float type flowmeters Brooks Instrument

Sho Rate R6-15-B (liquids)

Sho Rate R6-15-C (air)

Sho Rate R6-15-D (H2S/N2)

USA

Float type flowmeter

in Teflon® Gilmont GF-1360 USA

Centrifugal pumps Iwaki Magnet Pump MD-30FX Japan

Gas membrane pump KNF N816.1.2KN.45.18 France

Static mixer Koflo 3/8 OD Japan

Bourdon manometers Bamo 63 P60.5 France

Ozone generator Trailigaz 24 g h-1

fed with O2 France

Ion chromatograph Dionex DX 120 USA

Ozone analyzer Trailigaz Uvozon 200 France

Combined pH-T probe Schott Blue Line 14 pH Germany

O2 probe WTW CellOx 325 Germany

Table A. 3: Results of the ozone absorption experiments using the operating conditions of Table 1.

CH2O2,i

(mg L-1

)

CH2O2,o

(mg L-1

)

pH QL

(L h-1

)

QG

(NL h-1

)

USG

(m s-1

)

EffO3

(%)

a°

(m2 m

-3)

Ha

1173 606 9.9 6.9 482 1.13 21.3 75.1 4.16

1173 538 9.9 6.9 606 1.27 20.8 80.2 3.92

1173 797 9.9 14.0 680 1.34 23.6 76.7 4.78

1173 762 9.9 12.2 762 1.44 21.5 77.7 4.67

1471 1071 10.04 14.0 642 1.18 33.2 76.1 5.93

1471 1197 9.99 21.2 720 1.27 33.6 81.3 6.11

1683 835 10.07 6.9 462 1.18 32.2 77.9 6.93

1683 888 10.07 6.9 496 1.25 28.5 70.0 6.73

1683 1250 10.03 14.0 567 1.36 30.8 71.4 6.91

1683 1262 10 14.0 633 1.44 29.1 72.8 7.56

1794 947 10.15 6.9 500 1.24 33.7 74.7 7.52

1794 1360 10.07 14.0 577 1.34 34.2 71.4 7.58

1794 1375 10.06 14.0 650 1.40 32.8 72.7 7.90

1794 1496 10.03 21.2 754 1.46 33.4 78.3 7.93

2720 1701 10.01 6.9 465 1.18 42.2 85.3 8.29

2720 1792 9.99 6.9 498 1.25 37.8 76.8 8.09

2720 2298 9.92 14.0 594 1.30 34.0 64.8 8.52

2720 2238 9.91 14.0 628 1.45 37.5 85.0 6.49

2720 2425 9.92 21.2 717 1.54 35.1 78.0 6.48

kinetic study of hydrogen sul de absorption in aqueous ... · water. chlorine is a top-grade...

Documents